← All Models

DeepSeek V4 vs Kimi K2.7 Code

DeepSeek V4 leads in 2 of 8 categories

Arena ELO

Tie

Cheaper

DeepSeek V4

Faster

Tie

Detailed Comparison

CategoryDeepSeek V4Kimi K2.7 Code
ProviderDeepSeek Moonshot AI
Context Window128K 256K WIN
Input Price ($/1M)$0.14 WIN$0.6
Output Price ($/1M)$0.28 WIN$2.4
Arena ELON/A N/A
CodingN/A N/A
ReasoningN/A N/A
Speed (tokens/sec)N/A N/A
Open SourceYes No

Frequently Asked Questions

Which is cheaper, DeepSeek V4 or Kimi K2.7 Code?

DeepSeek V4 is approximately 86% cheaper than Kimi K2.7 Code per million tokens.

Which has a larger context window?

DeepSeek V4 supports 128K context, while Kimi K2.7 Code supports 256K.

Is DeepSeek V4 or Kimi K2.7 Code open source?

DeepSeek V4 is open source. Kimi K2.7 Code is proprietary.

Related Articles

OpenAI’s distillation playbook explains the Kimi panic

Kimi K3 Is Forcing Silicon Valley to Pick Sides

Kimi K3 vs GLM-5.2: a one-endpoint test

Kimi K3 handles an 820k-line Rust codebase

Kimi K3 Benchmark Evaluation Guide for Coding Agents

Kimi K3 Proves Intelligence Still Costs Too Much

More Comparisons

DeepSeek V4 vs Claude Opus 4.5DeepSeek V4 vs Llama 4 ScoutDeepSeek V4 vs Claude Haiku 4.5DeepSeek V4 vs Qwen 3.6DeepSeek V4 vs Llama 4 MaverickDeepSeek V4 vs GPT-5.3-CodexDeepSeek V4 vs Gemma 4DeepSeek V4 vs GPT-4oKimi K2.7 Code vs Claude Opus 4.5Kimi K2.7 Code vs Llama 4 ScoutKimi K2.7 Code vs Claude Haiku 4.5Kimi K2.7 Code vs Qwen 3.6Kimi K2.7 Code vs Llama 4 MaverickKimi K2.7 Code vs GPT-5.3-CodexKimi K2.7 Code vs Gemma 4Kimi K2.7 Code vs GPT-4o